Wider community screening could sharply reduce tuberculosis burden

Global efforts to cut tuberculosis (TB) are not moving fast enough to meet international targets, and researchers are exploring new ways to speed up reductions in disease. Katherine C. Horton and colleagues used mathematical modelling to ask whether community screening could be part of the solution, and to weigh trade-offs between which diagnostic rules are used, how much of a population is screened, and how long screening campaigns last. The model separated people with TB into three kinds of disease: symptomatic infectious TB (people who reported symptoms at screening), asymptomatic infectious TB (infectious but not reporting symptoms), and non-infectious TB. By explicitly recognising these categories, the team could test screening strategies that focus only on people with prolonged cough, strategies that aim to find any infectious TB, and strategies that aim to find all TB disease. The study is framed around common, real-world diagnostic options and explores how screening intensity and duration change long-term effects on TB incidence and deaths.

To compare approaches the authors simulated three diagnostic algorithms: one targeting symptomatic infectious TB defined as prolonged cough with confirmatory Xpert Ultra; one targeting infectious TB using Xpert Ultra alone; and one targeting all TB using chest X-ray. They varied population coverage from low to maximal (100%) and considered up to five rounds of screening, projecting outcomes over a 10-year horizon. With maximum coverage and five rounds, the algorithm that targeted symptomatic TB was projected to reduce symptomatic TB incidence by 26.9% (22.8-31.5%), while the algorithm targeting infectious TB with Xpert Ultra gave a larger reduction of 74.0% (68.5-79.1%). However, incidence rebounded after screening stopped, erasing 9.8% and 15.9% respectively of those reductions within five years. The algorithm that aimed to detect all TB via chest X-ray offered the most rapid reductions—over 98%—with negligible rebound, but low diagnostic accuracy of current tools produced heavy overdiagnosis: an estimated 7.2 false positives per true positive in a single round. Results for TB mortality followed similar patterns.

These findings show that community screening can substantially reduce TB illness and deaths, but the choice of diagnostic approach matters a great deal. Algorithms that do not rely on symptoms—so-called symptom-agnostic approaches—tended to achieve greater impact with lower coverage and fewer rounds, because they can find infectious cases who would be missed by symptom-based screening. At the same time, tools that look for all TB with chest X-ray can deliver dramatic short-term reductions but create a large number of false positives given current diagnostic accuracy. The model also found that gains can fade after screening stops, highlighting the risk of rebound and the need for sustained strategies. To maximise and sustain epidemiological impact, the authors stress the need for better diagnostic tools and for treatment regimens that address non-infectious TB, so that early detection translates into long-term reductions in disease burden.